# Generative Search Policy Evaluation on RedNote Search (test) Source: [Generative Search Policy Evaluation on RedNote Search (test)]() Collection index: [llms.txt]() Data updated: 2026-03-12T08:44:00.920Z Evaluation of a generative search policy on the RedNote Search test set, assessing performance across query quality, evidence grounding, factual correctness, hallucination, formatting, content richness, and overall usability. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Generative Search Policy Evaluation | | Dataset (including any stated split) | RedNote Search (test) | | Metric | Query Quality Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Query Quality Score](): Higher is better; unit: PERCENT. - [Evidence Grounding Score](): Higher is better; unit: PERCENT. - [Basic Score](): Higher is better; unit: PERCENT. - [Hallucination Score](): Higher is better; unit: PERCENT. - [Formatting Score](): Higher is better; unit: PERCENT. - [Content Richness Score](): Higher is better; unit: PERCENT. - [Overall Usability Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Query Quality Score | Result date | Report paper | | --- | --- | --- | --- | | GRPO-Gated | 99.59 | Not recorded | 69b248a8d030084c22e2f9e1 | | RFT | 97.2 | Not recorded | 69b248a8d030084c22e2f9e1 | | GRPO-Linear | 96.36 | Not recorded | 69b248a8d030084c22e2f9e1 | | GRPO-GenRM | 96.3 | Not recorded | 69b248a8d030084c22e2f9e1 | | DPO | 95.4 | Not recorded | 69b248a8d030084c22e2f9e1 | | SFT | 91.67 | Not recorded | 69b248a8d030084c22e2f9e1 | ### GRPO-Gated Query Quality Score: 99.59. Result ID: 69b27c6f6dbb7075b95d266d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Optimization Strategy":"Group Relative Policy Optimization with Gated Aggregation"}. This record does not assert independent reproduction. ### RFT Query Quality Score: 97.2. Result ID: 69b27c6f6dbb7075b95d2669. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Optimization Strategy":"Rejection Fine-Tuning"}. This record does not assert independent reproduction. ### GRPO-Linear Query Quality Score: 96.36. Result ID: 69b27c6f6dbb7075b95d266c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Optimization Strategy":"Group Relative Policy Optimization with Weighted Sum Aggregation"}. This record does not assert independent reproduction. ### GRPO-GenRM Query Quality Score: 96.3. Result ID: 69b27c6f6dbb7075b95d266b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Optimization Strategy":"Group Relative Policy Optimization with Generative Reward Model"}. This record does not assert independent reproduction. ### DPO Query Quality Score: 95.4. Result ID: 69b27c6f6dbb7075b95d266a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Optimization Strategy":"Direct Preference Optimization"}. This record does not assert independent reproduction. ### SFT Query Quality Score: 91.67. Result ID: 69b27c6f6dbb7075b95d2668. [Reporting paper]() Recorded evaluation configuration: {"Optimization Strategy":"Supervised Fine-Tuning"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.